1 00:00:00,020 --> 00:00:04,210 Hi. I’m Cindy Starr. At NASA’s Scientific Visualization 2 00:00:04,210 --> 00:00:08,390 Studio we work with many Earth science 3 00:00:08,390 --> 00:00:12,570 datasets that must be accurately positioned on a globe. 4 00:00:12,570 --> 00:00:16,750 For example, we obtained 30-meter topography data for Greenland 5 00:00:16,750 --> 00:00:20,940 along with an ocean and an ice sheet mask from the Greenland Ice 6 00:00:20,940 --> 00:00:25,130 Mapping Project at Ohio State. Each of these datasets 7 00:00:25,130 --> 00:00:29,330 consists of a 6 x 6 array of 124 8 00:00:29,330 --> 00:00:33,510 megapixel tiles. We also received seven sets of 20-meter 9 00:00:33,510 --> 00:00:37,710 Radarsat data from the Canadian Space Agency, 10 00:00:37,710 --> 00:00:41,890 mosaicked at the University of Washington‘s Applied Physics Lab. 11 00:00:41,890 --> 00:00:46,080 Each set consist of a 5 x 5 array of 12 00:00:46,080 --> 00:00:50,260 421 megapixel tiles. We developed 13 00:00:50,260 --> 00:00:54,430 an IDL routine that extracts the coordinate and projection 14 00:00:54,430 --> 00:00:58,610 information from geo-tif images and writes out RSL 15 00:00:58,610 --> 00:01:02,830 include files that pass the parameters to a projection routine. 16 00:01:02,830 --> 00:01:07,080 The projection routine is a C++ Renderman plugin 17 00:01:07,080 --> 00:01:11,270 that computes the projection calculations in double precision 18 00:01:11,270 --> 00:01:15,480 and accurately positions the related texture tile. 19 00:01:15,480 --> 00:01:19,690 Parameters to the projection plugin are passed in string format in order to 20 00:01:19,690 --> 00:01:23,880 provide double precision accuracy. We employed this method with 21 00:01:23,880 --> 00:01:28,070 the high-resolution Greenland data, accurately mapping 87 22 00:01:28,070 --> 00:01:30,190 gigapixels of data.